A nonlinear-hybrid fuzzy/probabilistic load flow for radial distribution systems

Uncertainty analysis of future system operation is the key feature of current study. This paper shows that how subjective differences in representing fuzzy inputs can affect the final outputs. We consider a confined range of forecasted samples in terms of fuzzy membership functions as well as the range of historical forecasting errors in terms of probability density functions. On this basis, this paper proposes a combined fuzzy/probabilistic evaluation of distribution system voltages, considering hybrid fuzzy/probabilistic uncertainties for consumption and generation. The proposed method allows the sampled fuzzy inputs to incorporate in a Monte Carlo-based nonlinear fuzzy load flow including Wind Turbine Generating Units (WTGUs). As a result of hybrid uncertainty representation, the evaluation of fuzzy load flow results will be assessed in probabilistic framework. Numerical simulations are performed on the unbalanced IEEE 34-bus test distribution system equipped with WTGUs.

► The effect of error in modeling of fuzzy inputs on the fuzzy outputs is investigated.
► The error in fuzzification is regarded as PDF which forms a hybrid model.
► Unlike linear methods, the method uses a simple non-linear fuzzy load flow
► A Monte Carlo procedure runs the fuzzy load flow by sampling from hybrid inputs.